lca: Lexical Correspondence Analysis

In nicolarighetti/textools: an R toolbox for text mining tasks

Description

correspondence analysis of textual data

Usage

lca(
  df,
  docid_field = NULL,
  text_field = NULL,
  min_docfreq = 0.5,
  max_docfreq = 99
)

Arguments

df	a dataframe with at least a column with textual data and a column with documents' ID
docid_field	name of the column (in quotation marks) containing the IDs of the documents (default NULL)
text_field	name of the column (in quotation marks) containing textual data
min_docfreq	minimum values of a feature's document frequency, below which features will be removed (default 0.5 percentile)
max_docfreq	maximum values of a feature's document frequency, above which features will be removed (default 99 percentile)

Details

the function is substantially a wrapper of functions available in quanteda and factoextra. More specifically it leverages the correspondence analysis function CA. The fitted ca_model can be used to further explore the model and create plots with fviz_ca_biplot

Value

a correspondence analysis model

Examples

## Not run: 
dataframe$cluster <- lca(df, docid_field = "documents", text_field = "texts")
## End(Not run)

nicolarighetti/textools documentation built on Oct. 16, 2021, 11:20 p.m.